Locality and stratigraphy

Much of our knowledge regarding the composition of the original Burgess Shale biota comes from the Walcott Quarry (on Fossil Ridge) (Fig. 1), which has yielded tens of thousands of specimens and close to two hundred species [2], [11]. Less well known are several other Burgess Shale-type deposits discovered in the vicinity of the Walcott Quarry by the Royal Ontario Museum (ROM) since the 1980's [12]–[14]. Among those, the Tulip Beds locality (formerly known as the S7 locality [14]), discovered in 1983 on Mount Stephen (Fig. 1) [13] is probably one of the most important new finds. This site is so named because of the presence of the distinctive “tulip-like” fossils, which are both abundant and unique to this locality (with perhaps one exception from Utah; see below). In addition to numerous algae, the Tulip Beds material contains a low diversity, but relatively abundant, faunal assemblage. This includes sponges, lobopodians, priapulid worms, arthropods and several enigmatic forms [14]. These are currently being investigated as part of a larger community study project.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Geographical location of the Tulip Beds.

A, position of the site (yellow star next to arrow), in relation to the Walcott Quarry (green star) and the Trilobite Beds (blue star). (The GPS coordinates for the locality are 51°24′06.1″ N, 116°27′06.4″W). B, telephoto of the Tulip Beds from Mount Field, showing the location of the main fossiliferous outcrops (yellow stars). Fossils were collected from two talus slopes (yellow dashed lines) or in situ in the Above Camp outcrop (yellow star on the left), now represented by a small quarry. The yellow star on the right indicates the approximate extent of the Above Camp fossil bearing layers to the southeast (distance between the two outcrops approximately 200 meters). The white dashed line indicates the boundary between the Campsite Cliff Shale Member (CCS) above and the Yoho River Limestone Member (YRL) below. C, close-up view of the Above Camp quarry in the summer of 2010 (yellow star on the left in B), showing the location of the fossiliferous beds excavated by the Royal Ontario Museum (small quarry just above crew members).

https://doi.org/10.1371/journal.pone.0029233.g001

Stratigraphically the site lies within the lower section of the Campsite Cliff Shale Member, five to ten meters above the top of the Yoho River Limestone Member [13] (Fig. 1). This is considered to fall within the lowermost Bathyuriscus zone [14], Series 3, Stage 5 of the Middle Cambrian. Fossil specimens were collected by several ROM field crews from in situ outcrop just south of Fossil Gully Fault (Above Camp S7) and two talus slopes. Fossils from talus slopes were collected below the Above Camp S7 outcrop (AC), and along a slope 200–300 m to the southwest which samples the same fossil bearing interval (known as the S7 slope).

The Campsite Cliff Shale Member comprises blocky, silty mudstones, which are interpreted to have been deposited in deeper and more distal conditions than those of the closely associated, and slightly older, Trilobite Beds [13]. The fossil bearing lithofacies are comprised of calcareous claystones, characterized by mm-scale event deposited muds. The position of the Campsite Cliff Shale member in relation to the Cathedral Escarpment is unclear but the nearest evidence of the Escarpment is located in an outcrop on the northeast shoulder of Mount Stephen, about 1,200 meters from the Tulip Beds. A detailed stratigraphic and geological study of the Tulip Beds and its relationship to the Cathedral Escarpment and other Burgess Shale localities is beyond the scope of this paper.

Nomenclatural acts

The electronic version of this document does not represent a published work according to the International Code of Zoological Nomenclature (ICZN), and hence the nomenclatural acts contained in the electronic version are not available under that Code from the electronic edition. Therefore, a separate edition of this document was produced by a method that assures numerous identical and durable copies, and those copies were simultaneously obtainable (from the publication date noted on the first page of this article) for the purpose of providing a public and permanent scientific record, in accordance with Article 8.1 of the Code. The separate print-only edition is available on request from PLoS by sending a request to PLoS ONE, 185 Berry Street, Suite 3100, San Francisco, CA 94107, USA along with a check for $10 (to cover printing and postage) payable to “Public Library of Science”.

In addition, this published work and the nomenclatural acts it contains have been registered in ZooBank, the proposed online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix “http://zoobank.org/”. The LSID for this publication is: urn:lsid:zoobank.org:pub:94B075BE-B08D-4026-BF6F-F2D5383E1B54.

Taphonomy

Evidence of soft-deformation suggests that this animal was entirely non-mineralized. Like other Burgess Shale fossils [16], S. gregarium is preserved as compressed carbonaceous and aluminosilicate films parallel to bedding in a dark mudstone. All specimens were buried laterally or obliquely along the longitudinal axis (e.g. Fig. 7), except for one vertically compressed individual showing the radial axis (Fig. 8). The holdfast is occasionally preserved with a granular appearance at its base, which might suggest that it was once attached to coarser sediments (Fig. 9B,I,K). However, it is also possible that these granular elements represent differential tissue reactivity and partial decay of the holdfasts accompanied by early diagenetic mineralization. Examples of early diagenetic mineralization of cavities or tissues in other Burgess Shale species (e.g. midgut glands of Leanchoilia [17], gut diverticulae of Burgessia [16], eyes in Nectocaris [5]), may not be directly comparable, but at least, they suggest that such type of differential preservation is possible.

The quality of preservation varies greatly. It ranges from specimens with barely visible outlines or at least with some evidence of decay and dissociation (Fig. 6), to complete specimens with internal details preserved (Figs. 7, 8). While some variations could be attributed to diagenetic processes including post-burial decay, biostratinomic factors can be demonstrated in a number of specimens. For instance, elements of the calyx are sometimes disarticulated and found in close proximity with other parts of the same individual, including the stem (Fig. 6A, C). Such associations suggest pre-burial decay and possibly at most minimal transport after decay had started and during burial. The lack of significant transport is also suggested by the random orientations of specimens in clusters (Table 1, Fig. 15) and the lack or limited amount of sediment between specimens (Fig. 7). The fact that specimens within clusters are about the same size (Fig. 16) and tend to have similar preservation reinforces the view that entire populations were rapidly smothered and were not significantly transported.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 15. Rose diagrams showing the relative orientation of specimens of Siphusauctum gregarium n. gen and n. sp. on three different slabs.

A, ROM 61421. B, ROM 61428. C, ROM 61429. The measurements are relative as the original in-situ orientation is unknown. (See Table 1).

https://doi.org/10.1371/journal.pone.0029233.g015

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 16. Population variations of Siphusauctum gregarium n. gen. and n. sp within particular slabs.

Box and whisker plots: of total length (A) and calyx width (B). Grey indicates the entire data set, red indicates separate slabs with five or more individuals, and green indicates individual specimens and clusters of less than five specimens. The thick black lines indicate the mean value and the circles indicate outliers. Numbers below box and whisker plots refer to ROM numbers. N = 177.

https://doi.org/10.1371/journal.pone.0029233.g016

The external sheath decayed faster than the internal elements of the calyx. Only in specimens where decay is evident can the internal elements of the calyx (e.g. the feeding apparatus and the gut) be clearly seen (Figs. 3, 13). The loss of the external sheath can also result in internal elements appearing out of life position (e.g. comb-like segments splayed outwards, appearing twisted, or detached from the main body) (Figs. 6A, C, D, 13E).

The inner layer of the stem, the conical structure and the gut are the most decay-resistant parts of the animal. The stems are generally preserved straight or slightly bent (e.g. Figs. 5, 6A, 7, 9), have a uniform width, and have never been observed to be detached from the calyx. Like the external sheath of the calyx, the outer layer of the stem evidently decayed more rapidly so that in the most decayed specimens only the inner layer of the stems is preserved. The inner layer of the stem has noticeably higher relief than the outer layer (Fig. 9B, D), suggesting that it was more robust and composed of a tougher material than the outer layer. It is also possible that the inner layer was replaced by clay minerals during early diagenesis. The associated swellings around the outer layer might represent the accumulation of decay fluids (Figs. 6F, 9J). The conical structure connects the stem to the calyx and is always preserved with a similar shape (e.g. Figs. 5A, 11), suggesting it was more rigid and not flexible. The lowest part of the digestive tract is the most reflective part of the animal and appears black with a granular aspect in some specimens. Some specimens show evidence of sediment at the base and near the top of the calyx and just below the external sheath (e.g. Figs. 8B, 12D). This sediment is interpreted as the infilling of small holes in the external sheath probably connecting with the internal part of the calyx.

Among specimens of similar sizes, subtle variations in the shape of the calyx could be interpreted as resulting from a direct consequence of decay (resulting in the collapse of selective tissues) or alternatively from variations in the angles of burial. However, because there are no visible differences in preservation among specimens showing such variations in shape, it is not clear how decay alone could have had an impact and why it would have only affected some but not all specimens of the same size on the same bedding planes. Accordingly, differences in angles of burial among specimens might have played a more important role. Whereas angle of burial could have an influence on the length of the specimens and outline shape, it would be inconsequential for the maximal width, given that the stem and calyx are roughly circular in cross-section. Importantly, few specimens show deformations along the stem (i.e. compression artifacts which would be expected if the stems were buried at an angle) suggesting that most specimens were buried parallel to bedding. For specimens of similar widths and lengths, variations in the shape more likely represent a genuine feature of the animal and demonstrate the extent of flexibility of the stem and calyx encountered in the population (Fig. 10, Table 2).

Systematic Palaeontology

Unranked stem-group bilaterian

Phylum Uncertain

Family SIPHUSAUCTIDAE fam. nov.

Mode of life

The unusual morphology of this animal and the lack of extinct or extant analogues results in difficulties when establishing its mode of life. Despite these difficulties there are enough morphological clues to speculate on how it lived. The large size, stemmed nature, overall shape, and internal characters imply it was an active and semi-sessile benthic filter-feeder. The radially arranged comb-like elements are interpreted to function as a filter feeding structure. We suggest that expansion and contraction of the comb-like elements, and possibly the external sheath, created a pumping mechanism which allowed active water circulation through the calyx. Expansion of the calyx would result in water being drawn through the openings at the base and contraction would have resulted in the expulsion through openings at the top. It is possible that water might have exited through the openings at the base of the calyx as well. The fine striations at the surface of the comb-like elements could represent remnants of connective tissues belonging to muscle bands or less likely could be part of a filtration structure. Similar fine striae are present in the possible stem-group cephalopod Nectocaris (fig. 1d and suppl. fig. 12 in [5]) with a similar distance 0.1–0.15 mm between each striation. While Nectocaris is unrelated to Siphusauctum, these striae point to analogous structures.

Active pumping would require a network of circular muscles in the calyx of Siphusauctum and, although muscles themselves are not preserved in the Burgess Shale (e.g. [18]), it is conceivable that connective collagenous tissues which are less prone to decay could be preserved in both organisms. Presumably, the food would have passed from the transverse grooves to the six central grooves belonging to each of the comb-like elements, via the beating of putative cilia (not preserved). Food particles would have circulated down towards the gut, through a central mouth which has not been identified, but is suggested by the concentration of organic matter in this area. The digested food could have been expelled through the terminal anus and transported away without contaminating the inhalant flow of water.

Although the Ctenophora are not considered to be closely related to S. gregarium (see below), the apparent similarity between the comb rows of ctenophores and the comb-like segments of S. gregarium warrants functional comparison. Modern ctenophores have eight rows of equally spaced combs (plates of hundreds of fused cilia) found externally on the body that function in locomotion when beaten together synchronously. Primitive ctenophores have been recognized in both the Burgess Shale and Chengjiang faunas and differ from modern species by having up to 24 comb rows in some of the Burgess Shale forms and lacking tentacles [19]. The narrow and abundant comb rows of the Burgess Shale ctenophores are not morphologically comparable to S. gregarium. However, the putative Chengjiang ctenophore, Trigoides aclis, has four to eight broad comb rows [20] that are more similar to the comb-like segments in S. gregarium. Despite some similarity, the comb-like segments are not considered morphologically equivalent, because S. gregarium has segments that vary in both width and thickness and extend from either side of a longitudinal groove within the calyx. This type of variation is not seen in fossil or extant ctenophores. The presence of these structures internally rather than externally also suggests that they are not functionally comparable, and a locomotory function is not considered for the comb-like elements of S. gregarium.

Although S. gregarium lived in large groups, there is no evidence of any organic connection between individual specimens, suggesting they were not colonial (i.e. each specimen is a solitary individual). There is no indication of bulging in the stem or calyx which could indicate a particular type of propagation. The presence of a single specimen possibly comparable to S. gregarium in the Middle Cambrian of Utah (Lieberman, pers. comm. 2009), would suggest that members of this group occurred at least along the current Western Laurentian sea board during the Middle Cambrian.

Siphusauctum gregarium was probably well adapted to Proterozoic-style soft substrates [21]. The presence of fragile hexactinellid sponges like Diagoniella and large sponges [22] in the Tulip Beds, supports low sedimentation rate and the idea that the environment was relatively deep and not under the influence of strong currents. Also, given the presence of soft-sediment dwellers (e.g. priapulids), it is unlikely that the seafloor was very firm, and would not have provided a compact base for S. gregarium to attach. The lack of a large holdfast and variation in holdfast morphology suggest that S. gregarium lived in a very low current environment and/or was not permanently anchored. The holdfast would have been located within the flocculent upper mud layer, with the position of the outer layer on the stem giving the approximate level of the water-sediment interface (Fig. 17). We interpret the holdfast acting as a temporary anchor. It is possible that the holdfast could retract partially within the stem, as suggested by the presence of bulbous holdfasts with an extended outer layer. Such variation could have led the holdfast to detach itself from the substrate. Because the length of the stem varies greatly, it suggests that the inner layer of the stem might have been filled with a hydrostatic fluid; presumably allowing the holdfast to expand and contract.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 17. Reconstructions of Siphusauctum gregarium n. gen. and n. sp.

Left, outer appearance. Right, transparent view through the external sheath to show the internal anatomy of the calyx. (© Marianne Collins).

https://doi.org/10.1371/journal.pone.0029233.g017

As a facultative sessile, stalked filter feeder, the ecological strategy of S. gregarium has no direct counterparts with any other organisms, with the exception perhaps, of Dinomischus [23] which is also interpreted as a shallow sediment sticker [21], albeit with important differences (see below). Comparisons with other shallow sediment stickers, in particular from the Ediacaran, are tenuous. Among the forms within the rangeomorphs which have a holdfast and a stem, the structure of the frond differs fundamentally in being flat and elongate and having a fractal geometry (e.g. [24]). S. gregarium belongs to the highest tier (10 to 50 cm) in the Burgess Shale community and joins a few monaxonid sponges, Leptomitus and Wapkia which also occur in the Tulip Beds [22], and Mackenzia, a putative sessile cnidarian [25]. A marked difference between S. gregarium and the aforementioned species is that S. gregarium is abundant and certainly represented a significant biomass in its habitat, which contrasts with other Burgess Shale localities, such as the Walcott Quarry, where most of the sessile filter feeders occupy a low to middle tier [2].

Biological affinities

Among extinct and extant bilaterians, S. gregarium superficially resembles organisms from a range of phyla, embracing entoprocts to tunicates. However, the absence of convincing homologies with any of these phyla suggests no close affinities with S. gregarium.

Among the non-bilaterians, the oval shaped Burgess Shale ctenophore, Ctenorhabdotus [19], has a gross morphology that is similar to the calyx of S. gregarium. However, as discussed in the previous section, the ctenophore comb rows are not comparable. Therefore, in addition to lacking a stem with holdfast and a defined gut within a body cavity, these animals are not considered to have any close biological affinity with S. gregarium. A similar conclusion is reached with the enigmatic attached leaf-shaped animal Thaumaptilon [25] from the Burgess Shale. This animal may have affinities with some frondose Ediacaran forms [25] or represent a potential stem group cnidarian [26], but affinities to S. gregarium seem very remote. Priscansermarinus [27], another problematic stalked Burgess Shale animal originally thought to be a barnacle, is most likely unrelated to this group [28]. This animal has a small holdfast structure and a short deformable stolon. Like S. gregarium, this animal was gregarious, had a central visceral mass in the calyx and both animals lacked external tentacular elements. Although the calyx region appears incomparable with S. gregarium (Fig. 18E), a revision of Priscansermarinus is needed before more precise conclusions about the affinities of these two organisms can be formulated.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 18. Other stalked animals from the Burgess Shale.

A, Herpetogaster collinsi, primitive ambulacrarian (ROM 58037). B–D, Dinomischus isolatus, problematica. B, holotype (USNM 198735), C, paratype, showing a flatter base of the calyx (MCZ 1083), D, paratype showing long narrow stem and bracts pulled together at the top of the calyx (ROM 32573). E, Priscansermarinus barnetti, an enigmatic stalked animal with a thick stem (ROM 36064). F, Lyracystis radiata, a basal echinoderm with a long stem, cup shaped calyx (here surrounded by a halo of pyrite) and branchlets (ROM 57229). Scale bars: A–E = 5 mm, F = 30 mm. Abbreviations: Br - Bract, Bt - Branchlet, Ca - Calyx, Cp - Chitinous plate, Ph - Pharynx, Seg - Segment, St - Stem, Stom - Stomach, Ta - theca, Te - Tentacle, Td - Terminal disc.

https://doi.org/10.1371/journal.pone.0029233.g018

Within the deuterostomes, some crinoids have a bulbous calyx on a long stem; however, they have arms with pinnules attached to the aboral region and possess a distinctive stereom. Stem group echinoderms are also found in the Burgess Shale; e.g. Lyracystis, a tall suspension feeding echinoderm with three V-shaped arms [29], but a close affinity is not recognized (Fig. 18F). Tunicates have a sessile adult form, but their internal anatomy is incomparable with S. gregarium. There is no evidence of a pharynx, incurrent and excurrent siphons or oral tentacles in S. gregarium, and tunicates do not have any feeding structures that are similar to the striated elements in the comb-like segments. Phlogites, a chalice shaped animal with a stalk, cup-shaped calyx and five branched tentacles, has been compared with the lophophorates [30], entoprocts and the gnathiferans [31], but its position is probably closer to ambulacrarians [32]. Phlogites possesses branched tentacles and an anus on the side of the calyx [31], features that are not known in S. gregarium. Despite a similar shape of the stem and calyx, both animals are probably unrelated. Herpetogaster, another tentaculate animal from the Burgess Shale related to Phlogites [32], possesses a disc at the end of a flexible stolon which could have presumably functioned like a holdfast (Fig. 18A). The flexible stolon had an internal tube, potentially originally filled with some hydrostatic fluids which could have functioned for controlling the length of the stolon. Such feature is vaguely reminiscent of the inner layer of the stem in S. gregarium. However, like Phlogites, it is more likely that adaptation to a semi-sessile lifestyle for a bilaterian filter-feeder organism encompasses similar transformations, from the development of a similar anchoring mechanism, and presence of a flexible stolon, to the development of a cone shaped calyx. All these features are most likely convergent in various groups.

Dinomischus isolatus, another stalked animal from the Cambrian, whose biological affinities have yet to be conclusively resolved, may represent the closest relative to S. gregarium. D. isolatus is a solitary stalked filter feeder with a calyx sitting on a long stalk and surrounded by a ring of bracts [23] (Fig. 18B, C, D). A second species of Dinomischus (D. venustus) was described by Chen et al. [33] from the Chengjiang lagerstätte and an undescribed species has been reported from the Kaili biota [34]. Dinomischus venustus is morphologically similar to D. isolatus, but was described with a long rod-like extension from the top of the calyx. This feature was considered to represent a long anal pore [33]; however, Chen and Erdtmann [35] disagreed on the interpretation of this structure. Erdtmann argued that the structure is the stem, twisted under the calyx, especially because the extension is similarly preserved, and has the same thickness as, the stem. Chen argued for the anal pore interpretation, because the structure appears to be on the same plane as the calyx [35]. The alternative interpretation, with the stem folded on top of the calyx, is probably more parsimonious, especially because sediment separates the two structures, and the base of the stem is not visible on the fossils. Dinomischus shares with S. gregarium a long, narrow, stalk and a large stomach sac near the base of the calyx, which preserves as a mass of highly reflective black carbonaceous minerals. The stem in Dinomischus does not seem to be organized with an inner and outer layer. However, this could be taphonomic, with the outer layer not preserved, as can be seen in a number of S. gregarium specimens. Both organisms have a distinctive rigid conical structure that gives shape to the base of their calyx and acts as the base from which the other elements attach to. A ring of filter feeding elements encircles the calyx and encloses a putative mouth and anus in both animals. The bracts of Dinomischus appear to be short and stiff, whereas the segments of S. gregarium are much larger, with striated sheaths and feeding grooves. There is at least some indication of wrinkled internal structure in the bracts of Dinomischus that may be reminiscent of the striated elements of S. gregarium (fig. 5 in Chen et al. [33]). However, to date, these structures are poorly described, and it is unclear if they are comparable. It is also noteworthy that finer details of the bracts could have also been lost, especially since smaller features in small animals like Dinomischus are less likely to be preserved. Perhaps the most important differences between the two organisms are the number of the filter feeding elements, with six in S. gregarium and about 18 to 20 in Dinomischus, and the apparent lack of an external sheath enclosing the bracts of Dinomischus. Overall the similarities suggest the two organisms may be related. However, recognizing conclusive homologous characters between these two organisms may be difficult, and detailed studies of the available material of Dinomischus would be required to confirm or dismiss this hypothesis.

Amongst extant protostomes, comparisons with the entoprocts and ectoprocts are equally difficult to support. While previously it has been suggested that extant entoprocts are possibly miniaturized descendents of a Dinomischus-like animal [23], [33], it is difficult to make any comparisons with S. gregarium. Entoprocts and ectoprocts are two phyla of tiny stalked, sessile, solitary or colonial animals, but the morphology of their calyx is not comparable, and the affinities of these two groups remains contentious (see discussion in Caron et al. [32]).

Regardless of whether S. gregarium and Dinomischus are related, possession of a gut would suggest a grade of organization above sponges, cnidarians and ctenophores, possibly as a stem-bilaterian or within bilaterians. However, if radial organisms are in fact derived from a bilaterian ancestor [36], the presence of a gut itself might not be particularly informative. Like other fossil taxa [4], S. gregarium may elude the title Problematica, by collection of more specimens (potentially from various sites) with better morphological details preserved to resolve preservational artifacts and examination of the similarities with other animal groups rather than focusing on the differences. In the case of S. gregarium the problems of its biological affinity are not due to poor preservation or lack of specimens but rather the lack of phylogenetically informative characters, similar to the difficulties in resolving the phylogeny of the entoprocts. The lack of informative characters and the potential for so many sister group hypotheses for S. gregarium are probably due to its mode of life and the biological constrictions it imposes. The interpretations of mode of life made herein are hypotheses based upon the reconstruction of a large collection of animals and considerations for potential taphonomic variations as well as comparisons to extant taxa, however the very unusual arrangement within the calyx limit these comparisons. Discoveries of new specimens and species in other Burgess Shale-type deposits, in particular from the Chengjiang biota, and a detailed comparison with Dinomischus might help resolve the affinity of this animal in the future.